Reading of sacroiliac joints on plain pelvic radiographs: Agreement between clinical practice and trained central reading. Results of the DESIR-cohort.
Abstract
: Objective Investigating agreement on presence/absence of radiographic sacroiliitis between local rheumatologists/radiologists and central trained readers (external standard). Method Inflammatory back pain patients (IBP;≥3 months,<3 years) suggestive of axial spondyloarthritis (axSpA) were included in the DESIR-cohort. Baseline sacroiliac-joint-radiographs were read by two central readers (modified New York), adjudicated by a third reader in case of disagreement, yielding a positive or a negative result (central-reading). The same radiographs were also read by local radiologists/rheumatologists rated 'normal', 'doubtful sacroiliitis', 'obvious sacroiliitis' or 'SI-joint fusion' (local-reading); positivity defined as: 1) at least unilateral 'obvious sacroiliitis' 2) 'bilateral 'obvious sacroiliitis' or at least unilateral 'fusion''. Agreement and misclassifications between central readers and central-reading versus local-reading were calculated (kappas). Results Interreader agreement between the central readers was moderate (κ=0.54); 108/688 radiographs (15.7%) were adjudicated. According to local-reading (at least unilateral 'obvious sacroiliitis'), 183/688 patients (26.6%) had sacroiliitis; according to central-reading, 145/688 patients (21.1%). Agreement between local-reading and central-reading was also moderate (κ=0.55); 76/183 patients (41.5%) with at least unilateral 'obvious sacroiliitis' (positive by local-reading) and 32/109 patients (29.4%) with 'bilateral 'obvious sacroiliitis' or at least unilateral 'fusion'' (positive by local-reading) were rated 'negative' by central-reading; 38/505 patients (7.5%) and 68/579 (11.7%) respectively without sacroiliitis (negative by local-reading) were read 'positive' by central-reading. Conclusion In recent onset IBP-patients, both trained readers and local rheumatologists/radiologists agree only moderately in recognizing radiographic sacroiliitis. A significant proportion of locally recognized ankylosing spondylitis patients is not confirmed by central-reading (false-positive), while a small minority is false-negative, indicating the necessity of re-evaluating the role of radiographic sacroiliitis as diagnostic criterion for axSpA. © 2014 American College of Rheumatology.